16:03:51 RRSAgent has joined #simile 16:04:13 mickBass has joined #simile 16:04:18 won't be able to join on the phone, sorry 16:04:24 Agenda 16:04:24 1. Round table update from Pis 16:04:24 2. Logistics for arrival of new hires 16:04:24 3. Review project task list - see below 16:04:24 1. ARTSTOR AND OCW DATASETS 16:04:26 Progress update from Mark: 16:04:27 - Now possible transform OCW data just with XSLT, no need to use Perl. The 16:04:29 transform does not do everything the Perl transform did, but this avoids all 16:04:31 the team installing Perl on their machines. 16:04:35 - Artstor topic, geographic and subject fields are now organized 16:04:37 hierarchically where appropriate. 16:04:39 - New ANT automatic build script in CVS allows all team members to rebuild 16:04:41 datasets, avoids need to upload entire datasets to CVS, team members just 16:04:43 need to retrieve updated XSLT scripts from CVS. 16:04:45 - Added type information to OCW dataset. 16:04:47 To do: 16:04:49 - Adopt a common way of displaying names in Artstor and OCW. 16:04:51 - Need to change the way typing is done in OCW so it is more compatible with 16:04:53 the existing LOM schemas. 16:04:55 2. HAYSTACK / SIMILE 16:04:57 Progress update from Steve 16:04:59 Progress update from Andy on Joseki / Haystack integration 16:05:01 To do: 16:05:05 - Need to update simile.ad file so Haystack can display revised OCW dataset. 16:05:07 - Now hierarchical information has been added to datasets, Haystack needs to 16:05:09 process this in faceted browser. 16:05:11 3. CUSTOM BROWSER 16:05:13 Progress update from Mark: 16:05:15 - Can now display both Artstor and OCW data 16:05:17 - Can display hierarchical facets e.g. geographic, subject and topic 16:05:19 - Uploaded to CVS, Rob has been able to retrieve and run on both Linux and 16:05:21 Windows. 16:05:23 To do: 16:05:25 - Write interface so it is possible to switch between using Lucene for 16:05:27 queries and RDQL. 16:05:29 - Add text search boxes to facets when it is not possible to display all 16:05:33 facet values. 16:05:35 - Add paging to facets when it is not possible to display all facet values. 16:05:37 - Fix Ant / XSLT / MTXSLT / Saxon 7 / ENTITY / DOCTYPE bug. 16:05:39 - Need to fix Ant build script so tasks that call XSLT, unzip or Jena 16:05:41 Schemagen only rebuild when necessary. 16:05:43 - Add Jena persistant model support. 16:05:45 - Demonstrate inferencing between the two datasets. 16:05:47 - Display facet frequency. 16:05:49 - Allow both alphabetic sorting of facets and sorti 16:05:51 roll call: andy, steve, rob, mick on the phone 16:05:53 mackenzie joins 16:05:56 stefano on irc 16:05:57 eric sends his regrets 16:06:18 Hires update 16:06:47 Ryan Lee offer out, likely to start early next week (not definite) 16:07:16 I'm arriving in boston on Saturday night 16:07:29 will start operational on monday 16:07:41 Good progress on SIMILE funding from review 16:08:40 Mick in Cambridge 28-30 Jan 16:08:52 what happened to the hackaton? 16:09:25 demo critical milestone 16:09:42 (david karger joins) 16:09:55 (Stefano, we'll address you question in the next agenda item) 16:10:29 ok 16:10:31 next milestone to achieve demo with scale 16:10:57 sorry for not being able to call in 16:12:05 decided to stick with OCW & ARTstor despite lack of public accessibility 16:12:26 +1 16:12:38 OK for hires to start at CRL; office, access OK, will need kit (PC etc) 16:13:02 marbut_ has joined #simile 16:13:17 Mackenzie has the thumbnails for artstor 16:13:29 Mackenzie: But they will not let us make it publically available 16:13:48 SteveGarland: Isn't there no copyright on thumbnails? 16:14:03 Mackenzie: The Mellon Foundation, don't have that position, we are not the copyright owners in this case 16:14:16 I've written a letter to say we will not make them publically available 16:14:34 so I'm looking for an alternative corpus we can use later without such restrictions 16:15:15 David: It's been a slow week, alround, I'm working us through the list of haystack items 16:15:24 that we put together last week 16:15:45 logistics for new hires: 16:15:58 Offices and passes no prob for Stefano and Ryan (for Monday) 16:16:11 Kit (PC) is maybe issue 16:16:39 no prob for me, my laptop is with me and I don't need anything else 16:16:59 maybe I'll need a windows machine later on, but I can go virtualPC for now 16:17:41 mickBass: there are threads on three clients going: based on Haystack, Haystack-web, and the custom client 16:17:43 3 clients.. Haystack rich/Web and Mark B's Web client 16:18:09 (oops, sorry rob) 16:18:43 need to review each client and associated issue lists with core tech team 16:20:17 when each is at the stage they can view the dataset 16:21:05 next thursday PI call time, can walk through each client and issue list with tech team (many in Cambridge, some remote via netmeeting) 16:21:24 good for me 16:22:31 Steve G and Rob to sort out logistics for that (PCs etc) 16:22:50 ACTION: Steve G and Rob to sort out logistics for that (PCs etc) 16:23:30 dataset: Mark has enabled all SIMILE people to build dataset from CVS (with Java + Ant installed) 16:24:12 can build locally and XSLTs can be updated to fix probs in the RDF serialisations etc 16:24:35 (This is IPSSources CVS, not Haystack CVS) 16:25:49 Have an ant target that recognises if HAYSTACK_HOME is set, and can then copy data to Haystack, so we can keep separate CVS 16:26:36 ACTION: Mark to chase up IPSSources to get David K IPSSources access 16:27:36 Nice job mark! 16:29:11 Transforms updated to use subclass relationships for hierarchical controlled vocabs (e.g. China, province) 16:30:05 Type info added to OCW. Looking into getting Haystack to display that. 16:30:24 (Low-level type info, e.g. date, number etc. was previously missing) 16:30:49 Now have common stylesheet to canonicalise names in Artstor + OCW 16:31:15 Need to keep OCW in line with LOM schemas, however 16:31:25 mickBass has joined #simile 16:31:42 mick is back after proxy-server difficulty.. 16:33:14 where to put artstor thumbnails? 16:33:52 Need to think about long-term infrastructure needs... worry about that after demo 16:34:22 Haystack/SIMILE progress update 16:34:53 Haystack Web browser client: Good progress, should have something usable by next week 16:36:10 ACTION: Steve to leave a Haystack running, so people can try out it remotely 16:38:08 Haystack/Joseki progress update: seems to work OK. Fine-grained queries rather than a big GET, so won't work for large data sets 16:38:37 potential for cache on Haystack side 16:39:42 Need to fix simile.ad file in Haystack, updating OCW data uncovered problems 16:40:01 ACTION: Need to fix simile.ad file in Haystack, updating OCW data uncovered problems 16:40:22 ACTION: Need to find out if/how Haystack deals with sub-classing information 16:41:11 Probably don't need Haystack/Joseki integration for demo 16:43:19 Progress on thin Web client. Now can display OCW + ARTstor simultaneously, as well as hierarchical facets. In CVS, has been run on Linux + Windows, some problems to fix 16:43:32 (problems related to instal//build) 16:44:58 Issues: Need to add interface can use Lucene or RDQL; searchable/paging of facet values (900+ even in small dataset); ant/XSLT probs; support for Jena persistent models; demonstrate inferencing 16:45:14 also UI issues, sorting facets by alpha/frequency etc. 16:46:06 CVS Access progress update: Cleaned up a bit 16:46:28 Invitations re-issued to those who never logged in 16:46:44 ACTION: Invitations must be acted on within 7 days 16:47:37 Identifying corpus subset: Kevin has done some work. uses 'visulaising cultures course': USCDAA segment 16:48:36 Need user testing for Haystack + custom browser 16:48:45 Eric's suggestion of brownsauce+ RDFNavigator 16:53:48 simile.mit.edu: need a machine for infrastructure (issue lists, wiki etc.)... but simile.mit.edu is heavyweight for that, can probably find a lighter-weight machine for that. simile.mit.edu could be used for demo's where heavy-weight processing required 16:54:09 could run Haystack Web client on it, but tricky since you need the Haystack UI up and running, so might not work 16:54:48 I would like that machine to be linux/freebsd, would make it much easier to administer for me 16:54:50 Send someone to Cannes W3C tech plenary with SIMILE demo? 16:55:17 Steve volunteers :) 16:55:31 (plenary in March) Andy going anyway 16:55:53 will likely demo in first 1 or 2 days, need to get slot soon 16:58:20 RRSAgent, pointer? 16:58:20 See http://www.w3.org/2004/01/22-simile-irc#T16-58-20 16:59:06 RRSAgent, bye 16:59:06 I see 6 open action items: 16:59:06 ACTION: Steve G and Rob to sort out logistics for that (PCs etc) [1] 16:59:06 recorded in http://www.w3.org/2004/01/22-simile-irc#T16-22-50 16:59:06 ACTION: Mark to chase up IPSSources to get David K IPSSources access [2] 16:59:06 recorded in http://www.w3.org/2004/01/22-simile-irc#T16-26-36 16:59:06 ACTION: Steve to leave a Haystack running, so people can try out it remotely [3] 16:59:06 recorded in http://www.w3.org/2004/01/22-simile-irc#T16-36-10 16:59:06 ACTION: Need to fix simile.ad file in Haystack, updating OCW data uncovered problems [4] 16:59:06 recorded in http://www.w3.org/2004/01/22-simile-irc#T16-40-01 16:59:06 ACTION: Need to find out if/how Haystack deals with sub-classing information [5] 16:59:06 recorded in http://www.w3.org/2004/01/22-simile-irc#T16-40-22 16:59:06 ACTION: Invitations must be acted on within 7 days [6] 16:59:06 recorded in http://www.w3.org/2004/01/22-simile-irc#T16-46-44